Advances in tailoring the porosity of tannin-based carbon xerogels

*Corresponding author. E-mail: [email protected] (A. Arenillas) Tel. +34 985119090 Fax +34 985297662 Advances in tailoring the porosity of tannin-based carbon xerogels 1 Natalia Rey-Raap1, Andrzej Szczurek2, Vanessa Fierro2, Alain Celzard2, J. Angel 2 Menéndez1, Ana Arenillas1* 3 1 Instituto Nacional del Carbón, CSIC, Apdo 73, 33080 Oviedo, España 4 2 Institut Jean Lamour -UMR CNRS-Université de Lorraine n°7198, ENSTIB, 27 rue 5 Philippe Séguin, CS 60036, 88026 Épinal Cedex, France 6 7 ABSTRACT 8 Usually, carbon xerogels are obtained from resorcinol-formaldehyde organic gels. 9 However, more cost-effective and eco-friendly carbon xerogels can be synthesized by 10 using tannins instead of resorcinol, provided that a suitable surfactant is added to 11 prevent the collapse of the structure. The use of tannin, a natural phenolic compound 12 derived from wood, allows obtaining carbon xerogels with controlled porosity, as the 13 porous properties of these materials can be tailored by an appropriate choice of the 14 synthesis conditions. In this work, tannin-formaldehyde xerogels containing different 15 amounts of surfactant and formaldehyde were synthesized in order to evaluate their 16 effect on the porous structure and chemical composition. It was found that porosity and 17 density depend greatly on the amount of surfactant. The lowest density and highest 18 porosity values -0.34 g/cm3 and 78%, respectivelywere obtained by adding 10 wt. % 19 of surfactant. It was also found that S-doped carbon xerogels can be easily synthesized 20 due to the strong affinity between the carbon in the structure and the sulphur from the 21 surfactant. Furthermore, statistical analysis showed that there is interdependence 22 between the effect of formaldehyde and the surfactant, especially in the case of volume 23 and pore size. Hence, the choice of the appropriate surfactant-formaldehyde 24 2 concentration is essential for controlling the formation of the porous polymeric 25 structure. 26 27 Keywords: carbon gel, tannin, surfactant, controlled porosity 28 29 1 Introduction 30 Carbon gels are porous materials that are obtained by drying and carbonizing organic 31 polymer-based gels (Elkhatat and Al-Muhtaseb, 2011). The great advantage of these 32 materials is that, depending on the preparation method and the nature, concentration and 33 chemical composition of the precursors, the porous structure can be controlled (Matos et 34 al., 2006; Rey-Raap et al., 2014). Hence, their final properties can be tailored to fit the 35 requirement of a specific application. The versatility of carbon xerogels is the reason 36 why these materials are perfect candidates for various applications such as adsorption 37 (Job et al., 2005), hydrogen storage (Zubizarreta et al., 2009), electrode material for 38 supercapacitors or batteries (Braghiroli et al., 2015b; Roberts et al., 2014), catalyst 39 support (Yang et al., 2014), thermal insulators (Tannert et al., 2015), etc. 40 41 The most common method of synthesizing organic gels is by means of the 42 polymerization of resorcinol with formaldehyde in the presence of a solvent (Elkhatat 43 and Al-Muhtaseb, 2011). The organic gel obtained can be dried under supercritical 44 conditions, by freeze-drying process or by evaporation, which give rise to aerogels, 45 cryogels and xerogels, respectively (Rey-Raap et al., 2014). The preparation of aerogels 46 and cryogels involves cumbersome solvent exchange processes and complex drying 47 stages (Al-Muhtaseb and Ritter, 2003). Thus, evaporative drying is the quickest and 48 simplest method and therefore, yields the most cost-competitive materials. 49 3 50 Although resorcinol and formaldehyde are the most commonly employed reagents, they 51 are costly and harmful to the environment. Consequently, studies with less expensive 52 and greener precursors have been undertaken over the last few decades (Rey-Raap et al., 53 2015; Szczurek et al., 2014). All the hydroxylated benzenes tested to date contain 54 reactive hydroxyl groups which make them similar to resorcinol in the way they react 55 with aldehydes (Arbenz and Averous, 2014; Lochab et al., 2014). Of all the possible 56 reagents, the use of tannin, a natural counterpart to resorcinol, for the synthesis of 57 carbon gels has attracted a great deal of interest in recent years due to the characteristics 58 of this reagent, which is “eco-friendly”, cheap and non-toxic (Braghiroli et al., 2015a; 59 García et al., 2014). 60 61 As in the synthesis of resorcinol-formaldehyde carbon xerogels, the concentrations of 62 all the reagents employed to polymerize tannin with formaldehyde (i.e., tannin, 63 formaldehyde, solvent and catalyst), play an important role in tailoring the final 64 properties of these porous materials (Amaral-Labat et al., 2013b; Rey-Raap et al., 65 2014). These concentrations have been reported in the literature to be related to the pH, 66 the percentage of solids and the weight ratio of tannin to formaldehyde (Amaral-Labat 67 et al., 2013b; Szczurek et al., 2014). The pH has been one of the most studied variables 68 as it has a great influence on the final properties of tannin-formaldehyde carbon gels 69 over the entire range of porosity (Amaral-Labat et al., 2013a; Amaral-Labat et al., 2015; 70 Amaral-Labat et al., 2013b; Szczurek et al., 2011a; Szczurek et al., 2011b). Other 71 studies focused on the influence of the pH and percentage of solids, have reported that a 72 decrease in pH and percentage of solids leads to an increase in macroporosity, while 73 mesoporosity reaches its maximum value with an increase in the percentage of solids 74 4 (Amaral-Labat et al., 2013b). It should be pointed out that most of the studies reported 75 to date were performed on aerogels and cryogels (Amaral-Labat et al., 2013b; Szczurek 76 et al., 2011a; Szczurek et al., 2011b), since the synthesis of carbon xerogels with tannin 77 as precursor can only be achieved if surfactants are added (Amaral-Labat et al., 2013a; 78 Amaral-Labat et al., 2015). In a study on the synthesis of tannin-based xerogels, 79 Amaral-Labat et al. reported the effect that modifying the concentration of Pluronic F-80 127 from 5 to 20 wt. % has on the final porous properties (Amaral-Labat et al., 2015). 81 These authors found that xerogels turned from macroporous to mesoporous materials as 82 the amount of surfactant was increased. However, triblock copolymer surfactants such 83 as Pluronic require both an increase in temperature and the addition of an organic 84 solvent to dissolve and, hence, an extra step is required in the synthesis process 85 (Amaral-Labat et al., 2013a; Amaral-Labat et al., 2015). However, a recently published 86 work has demonstrated that tannin-based xerogels can be synthesized by using sodium 87 dodecyl sulphate (Rey-Raap et al., 2015). The use of this latter surfactant is expected to 88 give rise to S-doped materials. The introduction of heteroatoms during the synthesis 89 process is of great importance. For instance, the production of S-doped carbon materials 90 can enhance the properties of the started carbons and improve its suitability in some 91 applications concerning sorption and energy storage/conversion devices (Kiciński et al., 92 2014). 93 94 Despite the importance of the concentration of formaldehyde for tailoring the porous 95 properties, the effect of modifying the tannin to formaldehyde weight ratio from 0.5 to 96 0.6 has been evaluated, to the best of our knowledge, in only one study (Szczurek et al., 97 2011a), that focuses on carbon cryogels. Yet it is essential to appreciate the importance 98 5 of the concentration of formaldehyde for the polymerization reaction in order to control 99 porous properties of carbon xerogels. 100 101 The aim of the present work is to move a step closer towards the synthesis of tailored 102 porous carbon xerogels by a more cost-effective and environment-friendly route. To 103 achieve this goal, tannin-formaldehyde carbon xerogels were synthesized using different 104 amounts of formaldehyde and surfactant. The surfactant employed in the present study 105 does not require the addition of any organic solvent as it can be easily dissolved in 106 water. All the carbon xerogels synthesized were characterized from the point of view of 107 their chemical and porous structure in order to determine the effect of formaldehyde and 108 surfactant on the final porous properties of the xerogels. 109 110 2. Experimental 111 2.1 Raw materials 112 Commercial wattle tannin extracted from Acacia Mearnsii (De Wild) bark was used for 113 the present study. This raw material was kindly supplied by SilvaChimica (San Michele 114 Mondovi, Italy) under the name of Fintan OP. The industrial extraction process of 115 tannins has been explained in detail elsewhere (Braghiroli et al., 2014). Briefly, fresh 116 bark was subjected to leaching in a sodium bisulphite aqueous solution at 70ºC. The 117 resulting solutions were concentrated and then spray-dried to yield a light-brown 118 powder containing 80-82 wt. % of phenolic flavonoid materials, 4-6 wt. % of water, 1 119 wt. % of amino acids and other components such as monomeric and oligomeric 120 carbohydrates consisting of broken pieces of hemicellulose. The light-brown powder 121 thus obtained is referred to as tannin (T). 122 123 6 Formaldehyde (37 wt. % in water, stabilized by 10-15 wt. % of methanol), was supplied 124 by Roth. The sodium dodecyl sulfate and sodium hydroxide were purchased from 125 Sigma Aldrich and from Carlo Erba, respectively. 126 127 2.2 Preparation of the carbon xerogels 128 Organic gels were synthesized by the polycondensation of tannin (T) and formaldehyde 129 (F) using deionized water as solvent, sodium hydroxide as catalyst, and sodium dodecyl 130 sulfate (SDS) as surfactant (all the relevant data related to the reagents employed in this 131 study can be found in the Supporting Information). Tannin was first dissolved in 132 deionized water in open glass beakers under magnetic stirring at room temperature. 133 Once the tannin had dissolved, formaldehyde was introduced into the solution. Finally, 134 sodium dodecyl sulfate was added and the resulting mixture was stirred until a 135 homogeneous brown solution was obtained. All the gels were synthesized from 20 mL 136 of precursor solution with a pH value fixed at 5.5 by adding sodium hydroxide. The 137 tannin/formaldehyde weight ratio varied between 0.6 and 2.6 whereas the percentage of 138 solids remained fixed at 25 wt. % for all the samples. The concentration of sodium 139 dodecyl sulfate varied between 0 and 20 wt. %. 140 141 Each precursor solution was poured into a sealed glass tube, with an inner diameter 142 value of 1.4 cm, which was then placed in a ventilated oven at 85ºC for 72 h to undergo 143 gelation and curing. Afterwards, the tube was opened and introduced again into the 144 oven at 85ºC for 48 h for the organic gel to completely dry by evaporation of the 145 solvent. Once dry, each xerogel was carbonized at 900ºC under nitrogen in a horizontal 146 tubular furnace. The residence time was 2 h and the heating rate was set at 5ºC/min. The 147 7 resultant carbon xerogels were labelled ‘TF’ followed by the tannin-formaldehyde 148 weight ratio and the percentage of surfactant added, e.g., TF-1.2-10%. 149 150 2.3 Sample characterization 151 The porous properties of the carbon xerogels were characterized by means of mercury 152 porosimetry (Autopore IV, Micromeritics). Measurements were performed between 153 0.0035 and 414 MPa, and Washburn’s equation was applied to the corresponding data 154 to obtain the specific pore volume and average pore size (see supporting information for 155 more details). Skeletal and bulk densities were determined by helium (Accupyc II 1340, 156 Micromeritics) and envelope (Geopyc 1360 Envelope Density Analyzer, Micromeritics) 157 pycnometry, respectively. The overall porosity was determined from the bulk and 158 skeletal densities. The morphology of the carbon structure was observed using a Quanta 159 FEG 600 scanning electron microscope and the chemical structure was investigated 160 using an Infrared Affinity-1 spectrometer (Shimadzu, Japan) by applying an average of 161 20 scans at a maximum resolution of 4 cm-1 over a range of 4000–400 cm-1. For this 162 purpose, pellets were prepared by pressing 1 mg of sample ground together with 100 mg 163 of dry potassium bromide. The chemical composition of the samples was determined by 164 elemental analysis (C, H, O, N, S), using a Vario EL Cube microanalyzer (Elementar). 165 166 2.4 Statistical analysis 167 Response surface methodology was applied to the porous characterization data in order 168 to assess the significance of each of the variables studied (amount of anionic surfactant 169 (SDS) and tannin/formaldehyde (T/F) weight ratio) and of any possible synergy 170 between them. Significance was determined by applying an analysis of variance 171 (ANOVA) to each of the four porous properties selected as responses: porosity, bulk 172 8 density, pore volume and average pore size. The implementation of ANOVA allows the 173 interaction between variables and their effect on the main porous properties to be 174 evaluated with a minimum number of experiments. If the p-values obtained from the 175 ANOVA were smaller than 0.05, the corresponding variable would be assumed to have 176 a significant effect on the response. An optimal design covering all the possible SDS-177 T/F combinations together with repeated experimental runs were employed to obtain a 178 statistical model that adequately fitted the experimental results. The design matrix was 179 generated by using a Design-Expert 9.0 Trial version from Stat-Ease Inc. 180 181 3 Results and discussion 182 3.1 Statistical analysis 183 Response surface methodology was applied to the four variables selected as responses: 184 porosity, bulk density, pore volume and average pore size. Porosity and bulk density 185 were adjusted to a quadratic function, whereas pore volume and average pore size were 186 fitted to a cubic function. The R-squared values for porosity, bulk density, pore volume 187 and average pore size were 0.90, 0.93, 0.89 and 0.96, respectively, indicating that the 188 calculated results are in good agreement with those obtained experimentally. Moreover, 189 the data obtained by employing a least square technique indicate that the model applied 190 to each response is significant (the ANOVA data for each response variable are shown 191 in Tables S1, S2, S3 and S4 in the Supporting Information). The linear and quadratic 192 terms for the amount of surfactant have a significant effect on porosity and density, 193 whereas the volume and the average size of the pores depend on linear, quadratic and 194 cubic terms of the amount of surfactant, and also on the T/F weight ratio and the 195 interaction between the SDS and T/F weight ratio. Figure 1 shows the three-dimensional 196 surface plots constructed on the basis of model equations that represent the response 197 9 surface curves of the independent variables (T/F weight ratio and percentage of 198 surfactant) versus the dependent variables: porosity (Figure 1a), bulk density (Figure 199 1b), pore volume (Figure 1c) and average pore size (Figure 1d). From the ANOVA 200 analysis it can be inferred that the four variables studied are highly dependent on the 201 percentage of surfactant, while the T/F weight ratio significantly affects only on the 202 volume and average size of the pores. 203 204 Figure 1. Three-dimensional surfaces representing the effect of the T/F weight ratio and 205 percentage of surfactant on the porosity (a), bulk density (b), pore volume (c) and 206 average pore size (d) of tannin-based carbon xerogels. 207 208 3.2 Effect of the surfactant 209 In Figures 1a and 1b, clear differences can be observed when the percentage of 210 surfactant is modified. By adding 10 wt. % of surfactant, all the carbon xerogels exhibit 211 the lowest density and highest porosity values, 0.34 g/cm3 and 78%, respectively. An 212 increase in the percentage of surfactant to above 10 wt. % causes a slight decrease in 213 porosity from 78% to 70%. Conversely, a decrease to below 10 wt. % leads to materials 214 with a poorly developed porous structure, with an average porosity of less than 20% if 215 no surfactant is added. 216 217 The effect of the anionic surfactant (SDS) on the final porous structure of TF carbon 218 xerogels can be attributed to the polymerization of the tannin and formaldehyde. Since 219 reactive hydroxyl groups give tannin a reactivity towards formaldehyde similar to 220 resorcinol (Amaral-Labat et al., 2013a; Arbenz and Averous, 2014), it is possible to 221 deduce a mechanism for the reaction based on the well-known polymerization reaction 222 16 carbon xerogels synthesized have certain limitations for those applications requiring 373 narrow mesopores, but are totally useful in those applications in which larger pore sizes 374 are required. Furthermore, an additional advantage is that the surfactant used in the 375 present work allows sulphur to remain trapped inside the structure, which can be used as 376 a basis for obtaining S-doped carbons. 377 378 Acknowledgements 379 Financial support from the Ministerio de Economía y Competitividad of Spain 380 MINECO (under Projects MAT2011-23733, IPT-2012-0689-420000 and CTQ2013-381 49433-EXP) is greatly acknowledged. NRR is also grateful to MINECO for her 382 predoctoral research grant. The French authors also gratefully acknowledge the 383 financial support of the CPER 2007–2013 ‘‘Structuration du Poˆ le de Compe´ titivite´ 384 Fibres Grand’Est’’ (Competitiveness Fibre Cluster), through local (Conseil Général des 385 Vosges), regional (Région Lorraine), national (DRRT and FNADT) and European 386 (FEDER) funds. 387 388 References 389 Al-Muhtaseb, S.A., Ritter, J.A., 2003. Preparation and properties of resorcinol-390 formaldehyde organic and carbon gels. Adv. Mater. 15, 101-114. 391 Amaral-Labat, G., Grishechko, L.I., Fierro, V., Kuznetsov, B.N., Pizzi, A., Celzard, A., 392 2013a. Tannin-based xerogels with distinctive porous structures. Biomass Bioenerg. 56, 393 437-445. 394 Amaral-Labat, G., Szczurek, A., Fierro, V., Celzard, A., 2015. Unique bimodal carbon 395 xerogels from soft templating of tannin. Mater. Chem. Phys. 149–150, 193-201. 396 17 Amaral-Labat, G., Szczurek, A., Fierro, V., Pizzi, A., Celzard, A., 2013b. Systematic 397 studies of tannin–formaldehyde aerogels: preparation and properties. Science and 398 Technology of Advanced Materials 14, 015001. 399 Arbenz, A., Averous, L., 2014. Synthesis and characterization of fully biobased 400 aromatic polyols - oxybutylation of condensed tannins towards new macromolecular 401 architectures. RSC Adv. 4, 61564-61572. 402 Braghiroli, F.L., Fierro, V., Izquierdo, M.T., Parmentier, J., Pizzi, A., Celzard, A., 2014. 403 Kinetics of the hydrothermal treatment of tannin for producing carbonaceous 404 microspheres. Bioresour. Technol. 151, 271-277. 405 Braghiroli, F.L., Fierro, V., Izquierdo, M.T., Parmentier, J., Pizzi, A., Delmotte, L., 406 Fioux, P., Celzard, A., 2015a. High surface – Highly N-doped carbons from 407 hydrothermally treated tannin. Ind. Crops Prod. 66, 282-290. 408 Braghiroli, F.L., Fierro, V., Szczurek, A., Stein, N., Parmentier, J., Celzard, A., 2015b. 409 Electrochemical performances of hydrothermal tannin-based carbons doped with 410 nitrogen. Ind. Crops Prod. 70, 332-340. 411 Elkhatat, A.M., Al-Muhtaseb, S.A., 2011. Advances in tailoring resorcinol-412 formaldehyde organic and carbon gels. Adv. Mater. 23, 2887-2903. 413 García, D.E., Glasser, W.G., Pizzi, A., Lacoste, C., Laborie, M.P., 2014. Polyphenolic 414 resins prepared with maritime pine bark tannin and bulky-aldehydes. Ind. Crops Prod. 415 62, 84-93. 416 Job, N., Heinrichs, B., Ferauche, F., Noville, F., Marien, J., Pirard, J.-P., 2005. 417 Hydrodechlorination of 1,2-dichloroethane on Pd–Ag catalysts supported on tailored 418 texture carbon xerogels. Catal. Today 102–103, 234-241. 419 Kiciński, W., Szala, M., Bystrzejewski, M., 2014. Sulfur-doped porous carbons: 420 Synthesis and applications. Carbon 68, 1-32. 421 18 Lochab, B., Shukla, S., Varma, I.K., 2014. Naturally occurring phenolic sources: 422 monomers and polymers. RSC Adv. 4, 21712-21752. 423 Matos, I., Fernandes, S., Guerreiro, L., Barata, S., Ramos, A.M., Vital, J., Fonseca, 424 I.M., 2006. The effect of surfactants on the porosity of carbon xerogels. Micropor. 425 Mesopor. Mater. 92, 38-46. 426 Rey-Raap, N., Menéndez, J.A., Arenillas, A., 2014. Simultaneous adjustment of the 427 main chemical variables to fine-tune the porosity of carbon xerogels. Carbon 78, 490-428 499. 429 Rey-Raap, N., Szczurek, A., Fierro, V., Menéndez, J.A., Arenillas, A., Celzard, A., 430 2015. Towards a feasible and scalable production of bio-xerogels. J. Colloid Interface 431 Sci. 456, 138-144. 432 Roberts, A.D., Li, X., Zhang, H., 2014. Porous carbon spheres and monoliths: 433 morphology control, pore size tuning and their applications as Li-ion battery anode 434 materials. Chem. Soc. Rev. 43, 4341-4356. 435 Szczurek, A., Amaral-Labat, G., Fierro, V., Pizzi, A., Celzard, A., 2011a. The use of 436 tannin to prepare carbon gels. Part II. Carbon cryogels. Carbon 49, 2785-2794. 437 Szczurek, A., Amaral-Labat, G., Fierro, V., Pizzi, A., Masson, E., Celzard, A., 2011b. 438 The use of tannin to prepare carbon gels. Part I: Carbon aerogels. Carbon 49, 2773-439 2784. 440 Szczurek, A., Fierro, V., Pizzi, A., Stauber, M., Celzard, A., 2014. A new method for 441 preparing tannin-based foams. Ind. Crops Prod. 54, 40-53. 442 Tannert, R., Schwan, M., Ratke, L., 2015. Reduction of shrinkage and brittleness for 443 resorcinol-formaldehyde aerogels by means of a pH-controlled sol–gel process. J. 444 Supercrit. Fluids 106, 57-61. 445 19 Viana, R.B., da Silva, A.B.F., Pimentel, A.S., 2012. Infrared Spectroscopy of Anionic, 446 Cationic, and Zwitterionic Surfactants. Advances in Physical Chemistry 2012, 14. 447 Yang, T., Liu, J., Zhou, R., Chen, Z., Xu, H., Qiao, S.Z., Monteiro, M.J., 2014. N-doped 448 mesoporous carbon spheres as the oxygen reduction reaction catalysts. Journal of 449 Materials Chemistry A 2, 18139-18146. 450 Zubizarreta, L., Arenillas, A., Pis, J.J., 2009. Carbon materials for H2 storage. Int. J. 451 Hydrogen Energy 34, 4575-4581. 452 453